CV_Playbook_24

The issue of agent buy-in, or what NICE researchers call “the trust deficit,” is a top challenge for supervisors when it comes to contact center quality management, in general. According to the NICE survey results, 42 percent of respondents cited the likelihood of agents buying into the feedback they receive as the number one barrier to effective quality management, tied for first among the barriers listed with “the inconsistent application of quality management.” As it turns out, both of these challenges are directly related to the third challenge on this list: “random sampling that is not representative of agent performance.” If there is one take away from the NICE survey, it’s that contact center measurements and evaluations tend to rely on inadequate, skewed or random data. And that’s a big problem because that data is being used to make critical decisions. On average, as few as six digital interactions and 13 voice interactions per agent are being sampled every month by CX quality control programs. About a quarter of contact centers are measuring less than 10 voice and digital interactions combined each month, according to the survey data. Considering that all respondents work for contact centers with more than 200 agents, “this is statistically an insignificant sample size and not representative of agent performance,” stated the survey report. Meanwhile, 65 percent of respondents choose samples based on post interaction customer satisfaction surveys, which are known for attracting either highly satisfied or highly unsatisfied customers, potentially skewing the sample, argued NICE researchers. CSAT surveys also tend to have a low response rate, representing an insufficient sample of customers. Collecting CX data on digital interactions is particularly complex, and the technology is still maturing, which likely explains why 60 percent of companies don’t know what percentage of their customer digital interactions are being sampled, compared with just 9 percent for voice interactions. About half of companies target interactions based on speech analytics (55 percent). According to NICE executives, speech analytics tend to be used to check if the contact center staff followed a process or to make sure they said key phrases or didn’t use specific terms. This may be useful for checking processes but doesn’t necessarily drive improvements in customer satisfaction. In 51 percent of cases, control samples are chosen completely at random. Yet despite the lack of a statistically significant or holistic view, 85 percent of stakeholders said quality management programs influence business decisions across their organizations. That certainly sounds like a widespread problem. AI-driven quality management removes some of the subjectivity by automating the scoring of a wider array of agent soft skills across close to 100 percent of interactions, rather than just samples, providing a more holistic view of the customer experience, while at the same time reducing the influences of perceived human bias that create skepticism in scores. As with all computing breakthroughs, AI takes a once impossible, or at least impractical, level of data gathering and analysis and moves it toward the mundane. “As contact centers don’t have a truly holistic view of agent performance data, and are relying on small samples and inconsistent analysis, you cannot call decisions based on current quality management programs data driven,” said NICE executives. “Without being able to judge soft skills, organizations are viewing an incomplete picture.” The soft skills disconnect certainly is not a hidden problem or one that’s difficult to comprehend. After all, when it comes to assessing agent soft skills, 98 percent of respondents admit to challenges. And with the quality of the customer experience becoming a top area of focus among businesses the past few years, managing the quality of the call center is more critical than ever. Perhaps that’s why nearly all stakeholders surveyed for the NICE research report have plans to invest in AI analytics-driven quality management within the next 12 months or so, including the 14 percent that have done so already. o me employee work schedule? Manual processes QA team’s goals don’t align with the company’s goals Not measuring across all channels Lack of dedicated resources Random sampling is not representative of agent performance Sample size not representative of agent performance Agent buy-in to feedback Inconsistent application of quality management Top IoT Industries Based on Market Share CX leaders 41% 17% 31% 5% 6% ers CX leaders 38% 24% 9% 31% Source: NICE benchmark survey, 2023 Top Challenges of Effective Quality Management Source: NICE benchmark survey, 2023 42% 42% 38% 38% 36% 35% 31% 30% 10 THE CHANNEL MANAGER’S PLAYBOOK

RkJQdWJsaXNoZXIy NTg4Njc=